基于图像识别技术的鱼群浮头检测控制系统设计

Design of Detection and Control System for Fish Head Floating Based on Image Recognition Technology

摘要: 智能视频监控随着AI智能技术的发展，可以不需要人工干预就对检测目标进行识别、定位和跟踪操作。发展工业化养鱼离不开智能视频监控与智能管理的支撑，其智能监测的设计使工业化养鱼得到了发展。本文通过投喂平台的高清摄像头视频监控设备，监测并判别鱼群的浮动状态，结合NB-IOT模块和无线网桥，实现对鱼群集群行为、浮头行为进行实时监测的功能，同时在MATLAB的APP设计工具中，以图像识别理论为前提，将高清摄像头采集到的图像，结合图像处理算法，对鱼群图像进行边缘检测、平滑处理、边缘膨胀、数目标记等操作，以此对鱼群是否浮头进行判断，最后根据现有数据预测鱼群的行为走势。

Abstract: With the development of AI intelligent technology, intelligent video surveillance can identify, locate and track detection targets without manual intervention. The development of industrialized fish farming is inseparable from the support of intelligent video monitoring and intelligent management. Its intelligent monitoring design has enabled the development of industrialized fish farming. This article monitors and discriminates the floating state of fish schools through the HD camera video monitoring equipment of the feeding platform, combined with NB-IOT module and wireless bridge to realize the function of real-time monitoring of fish group behavior and floating head behavior. In the APP design tool, based on the premise of image recognition theory, the images collected by the high-definition camera are combined with image processing algorithms to perform edge detection, smoothing, edge expansion, number marking and other operations on the fish image to determine whether the fish are The floating head judges, and finally predicts the behavior of the fish school based on the existing data.